In hydrogenating organic materials using zerovalent metal catalysts, it is more common to use the metal dispersed on an inert support than to use, for example, colloidal dispersions of the metal itself. Included among advantages accruing to supported metals are their greater surface area, leading to increased reactivity, and their greater ease of separation, as by filtration. Colloidal metals are notoriously difficult to separate by filtration, and incomplete removal and recovery is costly and often deleterious to the product of hydrogenation.
When hydrogenations are conducted in aqueous media, the lack of hydrothermal stability of the commonly used supports places severe limitations on catalyst lifetime and recovery and also on the quality of the product due to dissolved support material. Where such hydrogenations are of hydroxylic organic compounds, the problem of hydrothermal instability of support materials is intensified. Where the organic compounds are polyhydroxylic, such as carbohydrates, the problem of hydrothermal instability is particularly exacerbated because of the relatively high concentration of hydroxyl groups from both water as solvent and the material to be hydrogenated.
The irony in hydrogenating aqueous solutions of carbohydrates is two-fold. First, the reduction products of many carbohydrates are important materials of commerce; sorbitol and mannitol are but two common reduction products. Second, there is no practical alternative to using water as the solvent in hydrogenating carbohydrates because carbohydrates generally are insoluble or, at best, sparingly soluble in most organic solvents. Because carbohydrates are solids, it is operationally mandatory to use a solvent in their hydrogenation.
I have previously disclosed a general method of hydrogenating carbohydrates in aqueous solution using as a catalyst zerovalent ruthenium dispersed on a hydrothermally stable support. I now disclose the further discovery that the activity of the aforementioned catalyst can be substantially increased if the zerovalent ruthenium is prepared by reduction with hydrogen of a ruthenium compound at a lower temperature than is conventionally used. Additionally, still further increased activity is achieved when the ruthenium salt dispersed on the support is not calcined prior to its reduction, also contrary to the conventional method of preparation.